The cultivation of cells, especially eukaryotic cells or mammalian cells, constantly calls for the use of special culture media that make available to the cells the nutrient substances and growth substances that are required for efficient growth and for the production of the proteins that are desired. As a rule, serum or compounds that are derived from serum (e.g. bovine serum) are used as a component of the medium in this regard.
However, in the case of the use of serum or protein additives that are derived from human or animal sources in cell cultures, numerous problems exist, especially if the starting material for the preparation of a medicinal agent that is to be administered to humans is made available via the cell culture.
In the case of such serum preparations, therefore, the composition and quality already vary from batch to batch just because of the dissimilarity of the donor organisms for such preparations. This represents a considerable problem, especially for the standardization of cell production and in establishing standard growth conditions for such cells. However, intensive and constant quality control of the serum material that is used is required in every case. However, this is extremely time-consuming and cost intensive, especially in the case of such complex compositions as serum.
Moreover, such complex preparations contain a plurality of proteins that can act in a disruptive manner, especially within the context of the purification process for the recombinant protein that is to be recovered from the cell culture. This applies particularly to those proteins that are homologous with or similar to the protein that is to be recovered. Naturally, these problems are especially acute in the case of the recombinant recovery of serum proteins because the biogenic pendant in the medium that is used (e.g., bovine protein) can be removed reliably within the context of purification only via quite specific differential purification (e.g., with antibodies that are directed specifically only against the recombinant protein but not against the bovine protein (Björck, L., J. Immunol., 1988, Vol. 140, pp. 1194-1197; Nilson et al., J. Immunol Meth., 1993, 164, pp. 33-40).
Another issue in the use of serum or compounds which are derived from serum in the culture medium is the possibility of contamination by mycoplasma, viruses BSE agents, or disease-inducing agents that are as yet unknown.
The addition of serum components in order to guarantee adequate adhesion of the cells to their surfaces and to guarantee adequate production of the desired substances from the cells has, apart from a few exceptions, been previously regarded as indispensable precisely for the cultivation of cells on solid surfaces. Thus with the method that is described in WO 91/09935, for example, it has been possible to achieve a process for the serum-free and protein-free cultivation of the FSME virus/virus antigen by means of the serum-free and protein-free cultivation of surface-dependant permanent cells, preferably vero cells (see WO 96/15231). However these are not recombinant cells but, rather, host cells that are used for the production of virus antigen in a lytic process.
In contrast to this, the cells that are used preeminently for a recombinant preparation, for example CHO cells, are capable of adhering only to a limited extent. Thus, CHO cells that have been bred by conventional methods bind to both smooth and porous microcarriers only under serum-containing conditions (see U.S. Pat. No. 4,973,616; Cytotechnology 9 (1992), 247-253). However, if such cells are bred under serum-free conditions, they lose this property and do not adhere to smooth carriers, or they become detached with ease therefrom if other adhesion-promoting additions, such as e.g., fibronectin, insulin or transferrin, have not been provided in the medium. However these are also proteins that are derived from serum.
Alternatively to this, the cells can be bred using the suspension culture technique as well as e.g., using the batch process or using a continuous culture technique. Cultivation preferably takes place using the chemostat process (Ozturk S. S. et al., 1996, Abstr. Pap. Am. Chem. Soc., BIOT 164, Payne G. F. et al., in “Large Scale Cell Culture Technology,” 1987, ed. Lydersen B. K., Hauser publishers; pp. 206-212).
Kattinger H. et al (Advances Mol. Cell. Biology, 1996, 15A, 193-207) describe the long term cultivation of cells in protein-free medium, but these cells must be cultivated on carriers and do not leave alternatives as continuous culture techniques. It is stated that these cells only show long term stability when adhered to the surface of carriers because of reduced growth and, as a consequence, reduced demand for growth factors.
In addition, attempts have been made on several occasions in the prior art to adapt cells to a protein-free medium starting from serum-containing conditions. However, in the case of such adaptation, it has been found repeatedly that, compared to serum-containing conditions, the yield of expressed protein and the productivity of the recombinant cells are markedly reduced in the protein-free medium following adaptation (Appl. Microbiol. Biotechnol. 40 (1994), 691-658).
It has also been found that, in the case of a high cell density, the production of recombinant proteins is considerably restricted on occasions. During attempts to adapt the cells to protein-free or serum-free media, instability with reduced growth of the cells, which are used, is also found repeatedly so that cells with reduced expression are produced, or even nonproducing cells are produced, whereby these have a growth advantage, relative to the producing cells, in protein-free and serum-free media, and this leads to the fact that these overgrow the producing cells and then, finally, the entire culture now generates very low product yields.